Medical Endoscopic Image Segmentation with Multi-resolution Deformation

Abstract

Image segmentation is an essential technique in image analysis. In spite of issues in contour initialization, boundary concavities and high-level computation, active contour models, also known as snake, are a popular and successful method for segmentation among researchers. Segmentation process in snakes consists of calculation of energy and deformation of contour. In this paper, we present a new deformation method for active contour model, multi-resolution deformation algorithm ensuring powerful time reduction in convergence of the contour to target boundary.